MARINE ECOLOGY PROGRESS SERIES Vol. 61: 263-272, 1990 Published March 22 Mar. Ecol. Prog. Ser. 1 l Recruitment of Ascophyllum nodosum: wave action as a source of mortality*

Robert L. Vadas, Wesley A. Wright, Steven L. Miller"

Departrnent of Botany & Plant Palhology, University of , Orono, Maine 04469. USA

ABSTRACT: The brown alga Ascophyllum nodosuni (L.) Le Jolis is a dominant rocky intertidal organlsm throughout much of the North , yet its inabilily 10 colonize exposed or denuded shores is well recognized. Our experimental data show that wave action is a major source of mortality to recently settled zygotes. Artificially recniited zygotes consistently exhibited a Type IV survivorsh~p curve in the presence of moving water. As few as 10, bul olten only 1 reLalively low energy wave removed 85 to 99'% of recenlly setlled zygotes. increasing the setting time for attachment of zygotes (prior to distilrbance from water movemenl) had a positive effect on survival. However, survival was significantly lower at high densities, and decreased at long (24 h) setting times, probably as a result of bacleria 011 the surface of zygotes. Spatial refuges provided significant protection from gentle water movement but relatively little protectjon from waves. These data indicate that zygotes are maladapted lor attachment in moving water and suggest that water movement is the primary faclor conlrolliny recruitment and distributional patterns of A. nodosum. These and earlier observations on the long-term lack of colonizat~on of denuded shores suggest that successful recruitment is highly epjsodic on all but the most sheltered shores. Because of the widespread don~ina~~ceof A. nodosum, disturbance by waves or currents, and stochastic events may play major roles in structuring intertidal comrnunilies In the Northwest Allantic.

INTRODUCTION nell 1972. Grant 1977, Vadas 1979, Denny et al. 1985). Wave energy also causes indirect mortality to juvenile Wave action is generally considered an important and adult assemblages through movemenl oI algal Iactor in the distribution and abundance of intertidal fronds, logs, rocks and debris (Connell 1961, Dayton organisms. There are, however, Iew quantitative data 1971, Velimirov 1983. Sousa 1984). Considerably less is on the specific effects of waves on patterns of known about the effects of wave forces on juvenile abundance. Biological exposure scales describe the stages of marine organisms. Wave action, however, direct or indirect effects of waves (Ballantine 1961, clirectly reduces the density of newly recruiled barna- Dalby et al. 1978). but they do not characterize what cles (Ortega 1981). Few studies have analyzed wave aspects ol wave energy are important to the dynamics effects on recruiting or juvenile algae. Nonetheless, of a species. Attempts to quantify wave forces are such data may be essential for understanding the noteworthy (Jones & Demetropoulos 1968, Harger abundance patterns and dynamics of a species and 1970, Doty 1971), although they all have limitations community organization (Sebens & Lewis 1985). (Denny 198%). In general, little attention has been Along the northern Gulf of Maine the dominant paid to measuring abiotic forces on rocky shores inlertidal organism is the brown alga Ascophyllum (Palum bi 1984). nodosum (L.) Le Jolis (Keser et al. 1981, Topinka et al. Waves affect intertidal communities by damaging 1981, Vadas & Wright 1986). Despite its abundance on indivicluals and by displacing plants and ani~~ials, these and other North Atlantic shores, 2 striking pat- either individually or in groups (Schwenke 197 1, Con- terns are apparent: a marked decline or absence of plants (especially juveniles) on wave-exposed shores and a general lack of recruitment on all shores (Olt- * Maine Agricultural Experiment Station Publication No. manns 1889, devirville 1953, Printz 1956, Boney 1966, 1407 . * Present address: NURC-FDU, #40 Estate Castle Coakley, Baardseth 1970). Twenty years of observation at St. Croix, U.S. Virgin Islands 00820 Lamoine Maine, USA, revealed no successful coloniza-

(c)Inter-Research/Printsd in F. R. Germany 264 Mar. Ecol. Prog. Ser. 61: 263-272, 1990

tion by A. nodosum clespite its abundance at this and capacity of 21 Oh,determined by drying for 48 h, boiling adjacent sites (Vadas & Wright 1986). This particl~lar in distilled watcr for 2 h to remove air, soaking for 24 h, population has noticeably declined in cover during the weighing ancl drying to a constant weight at 60°C. last 5 to 10 yr. Several other sites in Maine show similar The surlaces of some pottery chips (Expt 4) were recruitment patterns, with few exceptions (Keser & modified to s~rnulate natural surfaces and provide Larson 1984). refuges for zygotes and germlings. One half of the Although Littorina littorea (L) grazes on or disturbs unfirecl clay surface was imprinted with a nylon mesh sporelings of Asco~~hyllumnodosun) (Sundene 1973, netting while the other half was imprinted with A50 Watson & Norton 1985, Patterson '1986), cage experi- grade sandpaper. This technique created 4 mi- ments designed to exclude this herbivore have pro- crohabitats: vertical and horizontal grooves, flat sur- duced only one natural recruit since 1972 (Vadas faces, and pits. Approximately 32 squares per chip (flat u~ipubl.).Moreover, attempts to recruit A. nodosum surfaces, 5 X 4 mm) were formed by the mesh. The artificially by pouring zygotes over numerous caged depth and width of the grooves ranged from 0.1 to natural surfaces have also failed to produce significant 0.2 mm and from 0.15 to 0.2 mm, respectively. The recruitment (Vadas & Wright 1986). This apparent grooves were designated vertical (parallel to water enigma, the dominance of A. nodosum and the concur- flow) and horizontal (perpendicular to water flow); sur- rent lack of visible recruitment on these and European vival was analyzed separately in each. The number and shores, raises the question of what controls coloniza- size of the squares and characteristics of the grooves tion? varied slightly because of differential stretching of the Here we examine recruitment and early survival nylon and differences in firing and cooling tem- (5 24 h) of zygotes in artificial and natural wave energy peratures. erlvironments as a function of setting or attachment time, Zygotes were seeded onto pottery chjps as follows density and surface texture. In particular, weaddress the Gametes of AscophyUum nodosum are reIe,ised natur- following questions: (1) Does wave action affect recruil- ally over a 4 to 6 wk period (April to June) 111 Maine. mentantl,if so, what controlsits effect? (2)Doessetting or Eggs and sperm were obtained by separately forcing attachment time affect recruitment? (3) Are refuges release (by drying) of rnale and female receptacles in important to recruitment and early survival? Our experi- shaded trays. Gamete release occurred in 1 to 2 h on mentsdemonstrate that wave action, and water motionin mild sunny days, but took considerably longer under general, is a major source of mortality to zygotes 01 cool cloudy conditions. Gametes were collected by Ascophyllun~ nodosuzn. Longer setting times and rinsing receptacles in beakers containing cool (l0 to refuges enhance survival but only in gentle flowing 15°C) seawater. To initiate lerlilization, gametes of waters.These results suggest that zygotesof A. nodosum each sex were diluted with seawater, co~nbinedto are mr11ddaptc.d to water nioven~cr~l,and that natural produce a dark turbid broth and stirred for 30 S. In the recr11itrnr:ntis r.pisorlir, vscept perhapson vtbrysheltered first experiment 15 min was allowed for fertilization. shores and among holdfasts in dense stands. Subsequenlly a minimum of 30 min was employed. Following fertilization the zygote suspension was further diluted and gently poured over chips in shallow METHODS pans. The chips were covered by l cm of the suspen- slon and left ~~ndisturbed.The pans were shaded and Pottery chips were used as a synthetic substrate for covered witt: ,~lurnlnumfoil to prevent warming and recruitment because of their favorable attachment enhance germinatinn. After the prescribed setting qualities, water holding potential and ease of removal time, the solulion was careiully siphoned from the pans for sampling (Hanic & Pringle 1978). The clay consisted to minimint disturbance. The chips were randomized, of a mixture of Tennessee Ball Clay ancl Kaolin. Clays gently removed from the pans, subjected to wave or containing lron oxides were avoided because they may water movement ancl then placed in moist, covered be toxic. Silica, nephaline, bentonite and whiting were pans until sampled. Zygotes were regularly misted added to the mixture; silica to lower the melting point, (fine spray) with cold seawater to prevent desiccation nephaline and bentonite as suspension agents, and until censusing was complete. whiting as a colori~lgagent. Upon firing the clay lurncd Zygotes in Expts 1, 2, and 3 were counted by first a flat white provid~nga reflective rather than a heat- estimating densities of entire chips as light, medium absorbing surface. Clay was rolled out in 1 cm thick and heavy, and then by random subsampling using slabs, cut into 5 cm X 5 cm squares and allowed to dry eyepiece grids. In Expt 4, zygotes were co~intedon 18 to a leathery consistency. Holes (0.5 cnl) were drilled squares (flats) on light density chips (< 20 per flat). On through the centers of the squares which were then medium density chips (21 to 200 per flat), zygotes were fired at 983°C (Cone 8).The chjps had a water-holding counted on G randomly selected llat squares. For high Vadas et al.: Recruitment of an ~ntertidalalga

density chips (> 200 per flat), 6 randomly selected flats lnvolved in the low survival at long setting tirnes in were subsampled using an eyepiece grid. The number pilot studies and Expt 2. Zygotes were subjected to 9 of zygotes in randomly selected horizontal and vertical setting intervals in 2 series: (a) 0.5, 1, 2, 3 and 4 h and grooves was estimated and placed into 7 categories: 0, (b) 6, 12, 16, and 24 h. Following settlement, 6 repli- 1-25, 26-50, 51-75, 76100, 101-125 and > 125 cated chips were placed in a reciprocal shaker bath at zygotes per groove. Zygotes on the pitted surface were 10°C for 1 min at 68 oscillations min-'. This produced a counted in 10 randomly selected pits. Comparisons flow velocity of 6 to 8 cm S-' (determined with a between and within grooves and pits were difficult heated-bead thermister flowmeter; Vogel 1981), which because they contained unequal areas and were not resulted in a 1.5 cm overwash of each chip. Although easy to quantify. As a result, data for all surfaces were this experiment simulated intertidal water movements, converted to percent survival of control chips. Controls acceleration forces off the sides of the tank could have in all experiments consisted of chips handled identi- increased dislodgment of zygotes. Controls for Series cally to experimenlal chips except for the specific treat- (a) and (b) were based on the 4 and 24 h setting times, ment. respectively, and were maintained in calm water at Expt 1 was designed to simulate the recent settle- 10°C. Controls for the low, medium and high density ment of zygotes In the field and to detcrmlne the treatments averaged, respectively, (mean f SD) impact of natural (LOIYenergy) waves on survival. To 53 + 9, 448 + 31 and 5378 f 1434 zygotes cm-2 for simulate the short period available for natural gamete Series (a) and 16 + 2, 174 f 41 and 3006 * 537 release and zygote attachment during a flooding tide, zygotes ctn-' for Series Ib) To test for diflerences zygotes were seeded onto chips and allowed to set between density treatments, analysis o( variance undisturbed for only 15 min. Chips were randomized (ANOVA), linear regression and analysis of covariance and exposed to waves either individually (Trial 1) or as (ANCOVAR) were employed (Sokal & Rohlf 1981). a group by mounting them on plywood strips (Trial 2). Log (n + 1)lransformations were used to eclualize the The individual chips or strips were hand-held, appres- variance. sed to the rock surface, and subjected to 6 (Trial 2) or 8 To test for the possibility that attachment of zygotes (Trial 1) different wave treatments at Pemaquid Pt (S(.[-, was influenced by environmental conditions (long set- Fig. 1). Colltrol chips were rnaintciined in air, but m1stc.d ting Limes, high zygote densities and bacterial growth), with seawater, for the duri~tion ot the longest wave bacteria adhering to zygotes were removed at 1, 4, 7, series (16) in each trial. This provided a conservative 11, 15 and 24 h with a sonicator (Branson 2200). Fjve estimate of control levels for chips exposed to fewer zygotes per setting time mrc,re individually isolaled and waves. Control densities averaged 1228 k 260 and sonicated for 30 S in stcbrilo seawater. The suspension

270 + 41zygotes cm-"(mean + SD), respectively, fur was plated on 0.1 O/O glucose and tryptone agar, incu- Trials 1 (n = 3) and 2 (n = 5). bated at 10°C for 24 h, and counted. Expt 2 was designed to simulate the effects of low Expt 4 was designed to test the influence of reiuges, energy waves on survival. The wavemaker was an 10 1 sethng times and natural water movement on zygote asy~n~netricalopen-topped plexiglas box (one side with survival at contrastiny exposures: Pemaql~id Point, an an obtuse angle). The box was mounted on cylindrical exposecl site, and Montsweag Bay, a sheltered pivots over a plexiglas tank connected to flowing sea- estuarine site (Fig. l). Zygotes from plants collected water. As the box filled it became unbalanced and near Pefl~dquid Pi were allowed to set on chips spilled the waler, creating a wave 2 to 5 cm high. mounted on boards. A multifactorial design consisting Zygotes were seeded onto pottery chips and allowed of 3 wave treatments (1,10 and 100 waves], 4 mi- to attach for 9 diffcl-cnt time intervals. After seeding, crohabitat types (flat, pitted and vertically and horizon- but beforc testing, experimental and control chips were tally grooved surfaces) and 3 setting times (1,2 and 4 h) maintained in seawaler in large shallow pans. At the was utilized and replicated 10 times per treatment. end of each setting period 12 chips (3 per lreatment) Because of the absence of waves at Montsweag Bay, were carefully removed, placed on a rack 35 cm fro111 zygotes were immersed at the edge of the water during the wave machine and exposed to a series of 1, 2, 4 or 8 a flooding tide fo~5 S, 2 and 20 min. These periods waves. Control densities for the 24 h treatments aver- were equivalent to the times required to complete the 3 aged 1865 f 43 zygotes cm-' [mean f SD, n = 3). wave treatments at Pemaquid Pt. Densities of control Expt 3 was designed to test the influence oI low, chips (100 wave treatment) at Pemaqi~id Pt averaged medium and high zygote densities and different setting 1577 + 838, 1715 k 592 and 860 + 634 (mean -+ SD, times on survivorship. This experiment was conducted n = 10) cm -2 for 1,2and 4 h setting times, respectively. in gently moving water to minimize the dominating Control densities at Montsweag Bay (20 min treatment) effect of waler movement on other variables. This averaged 1814 + 1190, 2062 + 860 and 3072 k 1063 series was also used to determine if artifacts werc (n = 30) for the 1, 2 and 4 h treatments. The data were 26G Mar. Ecol. Frog. Ser. 61: 263-272, 1990

Expt 2

Survival of zygoles, allowecl to set over a wide range of times ancl exposed lo ~rlifically generated waves, exhibited a Type [V survivorship curve except at the 3 and G h setting times. At the shortest (5 rnin to 1 h) and longest (12 to 24 h)selling limes, survival was reduced. Survival appeared to be enhancecl at intermediate set- ting times and after l wave was al~nostan order of magnitude higher than the other treatments (Fig. 3).

1~0.0 e A 10.0. 6 0-0 5 mln A-A 15 min 1.0.

0 B 0-0 30 rnln A-a 80 rnln Fig. l. Location of exposed (Pernaquid Pt) and sheltered Z (Montsweag Bay) experimental sites 8 1.0. r t

converted to percent (as noted above), transformed with an angular transformation, and analyzed with a 3-way ANOVA.

RESULTS

Expt 1

Zygotes set for 15 min on smooth pottery chips and exposed to natural wave action (20 to 50 cm high 11 waves) showed a classical Type IV survivorship curve 012 4 8 (Fig. 2). Approximately 90 O/o of the zygotes were NUMBER OF WAVES removed by the initial wave in both trials. Fig. 3. Ascophyllurn nodosum. Influence of setting time (A and B = short, C = intermediate, D = long) and waves on relative survivorship of zygotes (mean ? SD)in a wave tank. '0' waves denotes controls. Note: the y-axes are not the same for each figure

Expt3

Survival ol zyclotes exposed to low flow velocities in a shaker bath was positively conelated with setting time up to 3 or 4 h (Fig. 4). S~rviv~~lcrl zygotes at 3 to 4 h was equivalent to control treatments (absence of water movement). The 3 density curves in Series (a) fit a cubic NUMBEROF WAVES polynomial regression line with c!r[~ial slopcts but Fig. 2. Ascophyllunl nodosum. Relative survlvorship of zygotes unequal intercepts. At high density survival of zygotes (mean SD) recruited [or I5 nlin and exposed to natural k was significantly lower than at medium and light waves at Peniaquid Pt. The absence of error bars indicates that the standard deviation was within the syn~bol.'0' waves densities (p < 0.05). Survival at low and medium denotes cantrols densities for lonq~.~.setting times was equivalent to Vadas et al.: Recruitment of an intertidal algd

Controls Series a

Fig.4. Asrophyflun? nodosun!. Influence of den- s IF sit* and setting times on survivorship of zygotes Ing E Trials conducted under low water movement + -1.0 and in 2 series: (a) 0.5 to 4 h. and (b) 6 to 24 h. Vz Data were log (n + l) transformed; values 0 shown are means + 95% CI. L, M, H denote 3 controls for low, medium and high density after - 1.5 . , , ...... , 4 h (Serics a) and 24 h (Series b) LMHLMH0.5 1 Ssrles a Series b SmlNGTIME (h)

controls through 24 h (Fig. 4). However, at high density there was a significant decrease in survival at the 1 HR. SET longest setting time (p < 0.05). Bacteria isolated from zygotes after 15 h (setting time) in medium and high densily treatments (n = 5 plates, 6 25 A-A REFUC~SrnTS counts plate-') produced significantly (p < 0.05) more colonies (60.1 t 23.4 and 44.4 + 29.1, mean + SD, 0 zygote-', respectively) than similar isolates from 1014 !LF0 density cultures (13.9 f 5.2, mean + SD, colonies zygote-'). Bacteria isolated from shorter treatments 2 HR. SET (< 15 h) grewatsimilarrates regardless of zygoledensity.

Expt 4

Zygotes set for 1, 2 or 4 h on flats or refuges, and exposed to a series of natural waves (20 to 40 cm high) at Pemac[uicl Pt exhibiled Type IV survivorship curves. Regardless of treatment, most recruits were losl with the Iirst wave. Because there were no significant differ- ences (p < 0.05) between pits and grooves (one excep- 6011; Table 21, these categories werc c.olnbined ior clarity (Fig. 5). Survival was always lowest on the flats NUMBER OF WAVES and ranged from 7 to 16 % after 1 wave and from 1 to 5 % after 10 waves. Less than 0.2 % survived after 100 Fig. 5. Ascophyllurn nodnstrm. Effects of succ:csslve waves, refuges and setting times on relative suwlvorship of zygotes waves. Survival in pits and grooves was significantly (mean + 95 % Cl) at Pemaquid Pt. '0'waves denotes controls. higher than on flat surfaces for 1 and 10 waves. How- Wave amplitudes ranged from 20 to 40 cm during the ever, cohorts in these refuges also approached extinc- experimenl tion after 100 waves. Survival of zygotes exposedto a gently flooding tide At Pemaquid Pt all main effects, except setting time, in Montsweag Bay was substantially higl~erthan at and all interactions were significant, whereas at Mont- Pemaquid Pt. Survival was again lowest on the flats, sweag Bay all main effects and the setting time vs ranging from 65 to 100'?/u after 5 5 and from 59 to 82 % immersion time interaclion were significant (3-way and 27 to 45 % after 2 and 20 min, respectively (Fig. 6). ANOVA, p < 0.051. Because of the occurrence of sig- Conversely, mortal~tywas almost non-existent in the nificant interaction terms and the dominating influence pits and grooves alter 5 s and 2 min. After 20 min, of waves on survival, the influence of surface texture or sunrival in refuges ranged from 80 to 92% in the 1 and setting time upon survivorship were examined a1 each 2 h treatments and was 49Y0in the 4 h treatment. level of wave (or immersion) exposure (Tables 1 and 2). Mar. Ecol. Prog. Ser. 61: 263-272, 1990

2982). Survival on flat surfaces was significantly differ- ent from textured surfaces at each wave treatment and at both sites except for 100 waves at Pemaquid Pt. Significant differences among setting times occurred at 50- 0-0 FLATS &-A REFUGES Pernaquid Pt but varied in an unordered manner in all .---. 25. 1 wave treatments. This was caused by variation in wave 0 1 HR. SET E o*. forces, especially in single wave treatments. Such vari- 4 ation was averaged out in multiple wave trials. No significant differences were observed among setting LL 0 75- tirnes at the shorter immersion periods in Montsweay M V 50- Bay. However, after immersion for 20 min zygotes established for l and 2 h exhibited better survival 25. 2 HR. SET (p 0.05) than those set for 4 h. LE < g On E 1oo.b DISCUSSION

Our results indicate that water motion is a major 50 - I source of mortality for zygotes of Ascophyllum 25. 4 HR. SET nudosum. Laboratory and field experiments, under 0' 0' relatively benign wave conditions, show that water 0 5 sec 2 rnin 20 min on (1) (10) (1 00) movement reduces recruitment protected and IMMERSIONTIME ('= NO. WAVES) exposed shores. Irrespective of setting time and whether or not refuges are provided, 85 to 99 % of the Fig.6. Ascophyllum nodosuln. Effects of immersion times zygotes are dislodged by the first 10 waves with most (= No. waves at Pemaquid Pt), refuges and setting times on relative survivorship of zygotes (mean 95% CI) at Mont- mortality occurring after a single wave. The severity of sweag Bay. '0' waves denotes controls. Water ~novement was this disturbance is further accentuated by the virtual limited to that caused by a gently floodjng tide loss of entire cohorts within 20 min. Moreover, these estimates of mortality are probably conservative Significant differences within each level of wave expo- because under natural tidal and water movements the sure were determined by the Ryan-Elinot-Gabriel- long undisturbed pe~iods required for initial attach- Welsch multiple F test (REGWF) (SAS Institute Inc. ment are uncommon. Gametes are usually released at

Table 1. Ascophyllurn nodosurn.Influence of surface texture on survivorship of zygotes at dicferent exposures

Surfaces Pemaquid Point (Exposed) Montsweag Say (Sheltered) Mean % survrvorship' Mean "/o survivorship '

l wave Horizontal grooves 31.3a Vertical grooves 31.4a Pits 38.9a Flats 9.9b 10 waves 2 min Horizontal grooves Vert~calyrooves Pits Flats 100 waves 20 min Horizontal grooves Vertical grooves Pits Flats

ANOVA with REGWF test on arcsine transformed data; p = 0.05, n = 30. [n each set of 4 entries, vdlues followed by different letters are significantly different. Vadas et al.: Recruitment of an intertidal alga 269

Table 2. Ascophyllum nodosum. Influence of setting time on survivorship of zygotes a1 diflerpnt exposures

Setting time Penlaquid Point (Exposecl) Montswcay Bay (Sheltered) - - Mean '%, s~rrvivorship' Mean 'X,survivorship' 1 wave BS.9a 19.5~ 26.5b 10 waves 2 min 3.7~ 88.9 12.0~1 93.4 ns 5.7b tlY li 100 waves 20 rnin O.lb 73.5a 0.7a 79.8~1 0.2b 44.5b

ANOVA with REGWF test on arcsine transformed data; p = 0.05, n = 40. In each set01 3 ~ntries,values followed by different letters are significantly different

ebb tide, especially on warm sunny days ancl are val was pos~tlvely correlated with setting time. How- washed free of the receptacles with the incoming tide ever, this relationship was confounded by high zygole (Bacon & Vadas unpubl.). Thus, little (motionless) time densities, especially at longer setting tinies which we is available for in situ fertilization and attachment. infer enhanced bacterial growth. Thus it is possible that Hence Expt 1, which shows an extreme Type IV sur- bacteria were responsible for the reduced survival at vivorship curve, most closely simulates the conditions long setting times in Expl 2. found in nature on moderate to wave-exposed shores. The inabillly of zygotes of Ascophyllui~~nodos~lm to These data also indicate that the movement of a flood- adhere cluickly and firmly to solid substrata is surpris- ing tide, excepl perhaps on highly sheltered shores, ing. Most algae produce mucilage for attachment e.g. Monlsweag Bay, makes attachment ul newly (Fletcher 1976), bul strategies in fucoicl algae vary formed zygotes extrenlely clifficult. consiclerably. Zygotes and ger~nlings of Sargassum Survival patterns of zygotes were similar in all exper- 111itlic.urn are sticky, sink ~[uicklyand attach rapidly iments despite the wide range of wave treatments and (Deyslier & Norton 1982). Several fucoids have delayed setting times used. Tlic shortest setting times (up 10 rhizoidal development and have specialized atlach- 1.5 h) were consistently least conducive to SUI-vival. nlent mechanisms, e.g. mesochiton in Pelvetia and spe- The reduced sur.vival of zygotes in the 4 h treatment at cial cell walls in zygotes of Bifurcariaand Hjmantlialia Montsweag Bay (Expl4) and the enhanced survival at (Hardy & Moss 1979, Moss 1981). Zygotes of A. nodo- intermediate selling times (Expt 2) should be noted. sum and spp. apparently lack stickiness but 'l'he 4 h treatment was run 16 d after the 1 and 2 h secrete mucilage through the wall (Moss 1981). Moss treatments, and near the end of the natural release (1975) showed that fertilized eggs of A. nodosum pro- p~riodfor Ascophyllum nodosum in Maine (zygotes duce a mucilage pad immediately after settlement formed from late season gametes appear to be dis- which attaches the zygote to the surface. Subsequently lodged morr easily: Vadas unpubl.). Preliminary a prirna~yrhizoid is produced which secures the gerrn- studies and Expt 2 indicated thal zygotes recruited for Ling to the substratum. Under culture conditions the shortest and longest intervals had the highest mor- rhizoids form within 24 h (Patterson 1986) and are tality: intermtldiate times (2 to 6 h) seemingly provided enhanced in darkness (Sheader & Moss 1975). The best survival. However earlier studies indicated that cluestion of why zygotes of A. nodosum do not altach survival should increase with setting time. Chartus et firmly rc5rnclinsunanswered. al. (1972) showed that carposporcs of red c~lgcic Althouyli 2 types of spatial refuges (pits and grooves) attached for 19 h were less easily dslodged Lhctri those were ulilized in our studies, neither was effectjvr in recruited for 7 to 9 h. Similarly, adhesion of germlings moving water. Microhabitat heterogeneity is often con- of Sargassum rnut,rum was enhanced with time sldelcd crilical to the survival of intertidal organisms (Norton 1983). (Bergeron & Bourget 1986). Some species settle in Expt 3 clearly demonstrated that attmhment or survi- refuges, and gain a foothold on exposed shores, escape 270 Mar. Ecnl. Prr~g.Ser. 61: 263-272, IU?)O

predation, or reduce desiccation and other stresses term observations suggest that large new patches are (Dayton 1971). Refuges for fucoid algae generally have unlikely to develop, except from synchronously been considered escape mechanisms from herbivorous recruit~dcohorts under calm sea conditions lasting for gastropods (Lubchenco 1978, 1983, Schonheck & Nor- several d~ys.Although modificatjon of wave energy by ton 1980. Vadas (,I al. 1982) which are thought to play a adult fucoid stands is also possible, similar to kelp beds significant role in mortality of juvenile fucoids (Sun- (Jackson & Winant 1983, Tecjner 19861, tidally induced tlvne 1973, Vadas et al. 1977, Keser et al. 1981, Lub- water movements probably eliminate all but a trickle of chenco 1980, 1986). However, refuges may also zygotes from recruiting into the understory (Lubchenco minimize other disturbances to fucoids. In ice-scoured 1986, P. Aberg pers. comm., Vadas unpubl.). Episodic regions of the North Atlantic Ascophyllum nodosum is recruitnient, which is well recognized in many organ- restricted to crevices and ice-free boulder fields (R. isms (White 1985) including marine invertebrates (Coe Wilce pers. comm.). In addition, desiccation, light 1956, Vahl 1982, Sebens & Lewis 1985, Tegner 1986) intensity and wave shock pressures (Carstens 1968, and recently algae (Reed et al. 1988),could account for Denny 19i35b) are reduced in crevices and mi- the presence of A. nodosr~mon moderate and exposed crohabitats. Survival ol germlings of Sargassum shores. Its presence and repopulation on highly pro- nluticum, for example, is higher in the lee of roughened tected shores may be more straightforward (Keser & surfaces (Norton 1983). Our experiments suggest that Larson 1984) although specific mechanisms al-e lacking. refuges enhance survivorship of A. nodosirnt in low The ability of Ascophyllun~ nuciosttm to persist as and moderate energy environments by ameliorating monocultures on many shores probably results from its disturbance from water movemetlt. This hypothesis is perennial habit and the presence of a dense understory supported by long-term {I to 2 yr] survival of germlings of suppressed shoots. Cousens (1986) and Vadas & in refuges beneath dense canopies on moderately Wrlght (1986) have argued that the presence of small exposed shores (Vadas, Miller & Wright unpubl.). relatively unbranched shoots or plants ('juveniles') is the The inability of a photosynthetic benthic organism to result of competition for light in dense stands rather than attach readily to natural surfaces cannot be reconciled persistent recruitment. These suppressed shoots Iunc- easily given the dynanlic nature of intertidal environ- tion as 'meristem banks' (Cousens 1985, Vodas & Wright ments. hfe in the Littoral zone entails coping with water 1986) and are thought to re-establish the bed following motion, presuniably Lhrough evolution. However, disturbance to the canopy. Once established these Ascophyllum nodosum appears to be maladapted to dense stands or patches of A, nodosum appear to occupy moving water, both in its juvenile stage, as shown here, space [or decades, and thus pre-empt space (Lewis and possibly in its adult form where high breakage of 1976) or oulcornpete other fucoids for space through shoots occurs regularly on exposed and other shares growth and longevity (Schonbeck & Norton 1980, Keser (Vadas 1972, Vadas et al. 1976, 1978). Interestingly, & Larson 1984, Cousens 1985, Vadas & Wright 1986). McEachreon & Thomas (1987) showed that breaking The absence of Ascophyllum nodosum from exposed stress in A. nodosum was positively correlaled with shores has long been recognized but not explained wave exposure, which suggests that adults are physi- (Lewis 1964). Lewis (1968) suggested that dispersal may ologically or genetically adapted to different degrees of be a problem and noted that sporelings were absent in wave exposure. It Inay be unreasonable, however, to exposed situations. Our inability to recruit zygotes expect all Life history stages of intertidal oryanisrns to poured over natural surfaces (Vadas & Wright 1986) have adapted to water movelnent. Sea anenomes on indicates that dispersal per se is not the cause of dis- exposed shores, for example, avoid mainstream vel- tribution patterns. similarly, the absence nf A. nodosurn ocities (Koehl 1977) and repopulate rocky shores from current swept areas (Lewis 1968, Mathieson et al. through asexual means (Sebens 1982), thus eliminaling 3977, Hardwick-Witman & Mathieson 1983) is probably 'the need' for specific adaptations by the zygote for related to attachment rather than dispersal. dealing with high wave forces. This stucly provides adhtional support for the poten- With Ascophyllunl nodosurn the inability to recruit tial importance of recruitment in structuring benthic readily may be compensated by iteroparity and an assemblages (Underwood & Denley 1984, Games & enormous annual investment in reproduction. A. Roughgarden 1985). Our data lndicale that wave action no do sum^is a perennidl (Baardseth 1970) and, before is an important lactor controlling the distribution and

I-cleasinggametes in the spring invests ca 40 to 50 '/Q of abundance of juvenile stages of a dominant species. its biomass in rsprod~.~ctivestructures (Josselyn & Although abiotic factors have been neglected in recent Mathieson 1978, Vadas unpubl.). Nonetheless, success- decades, they are as important as biotjc interactions in ful recruitment on exposed and moderately exposed determining patterns of abundance (Lewis 1980, shores probably involves a rare combination of mechan- Denny et al. 1985, Sebens & Lewis 1985, Underwood isms or stochastic events. Our experiments and long- 1986). Since Ascophyllum nodosum is the dominant life Vadas et al.: Recruitment of an intert~dalalga 27 1

form on many North Atlantic shores, factors aflecting Denny, M. W., Ihniel. T. L., Koehl, M. A. R. (1985). Mechani- its distribution and abundance may have important cal limits to size in wave-swept organisms. Ecol. Monogr. 55: 69-102 ramifications for the structure and organization of devirville, D. (1953). Depeuplenlent dc Id flor~marine sur les rocky intertidal communities. cbtes occidentales du Cotentin. Prt~c.Int. Symp. 1: 26-28 Ackno~vlcdgmlents. We thank Linda Bacon, Chris Bolis and Deysher. L.. Norton, T. A. (1982). Dispersnl and colonization in Patr~cidI'atlerson for field assistdncc., Drs Roger Cousens and Saryassum n~uticlul~(Yendo) Fensholt. J. exp. mar. Biol. tan Ilavison for review of earlier dralts and Jean Ketrh for Ecol. 56: 179-195 typing thc Ins. We appreciate the constructive criticism of John Doty, M. S. (1971). Measurement of water movement in refer- Pedrse ant1 an anonymous reviewer. We acknowledge the ence to benthic algal growth. Botanica mar. 14: 32-35 support of the Maine Agricultural Experinlent Station, the Fletcher, R. L. (1976). Post-germination attdchment mechan- NOAA. National Sea Grant Program to the University of Maine isms in marine fouling algae. In: Sharpley, J. M., Kaplan, (NA-80-AA-D-00034) and a faculty equipment grant from Lhe A. M. (eds.) Proc. Third Int. Symp. on Biodegradation. University of Maine. R. L. V. acknowledges the Department of Applied Science Ltd., London, p. 443464 Marine Zoology, University of Oslo, The Norwegian Institute of Gaines. S., Roughgarden, J. (1985). Larval settlement rate: a Water Resources (NIVA) and the Norwegian government leading determinant of structure in an ecological commun- (NTNF) fox providing space and support to write. ity of the marine . Proc, nall Acarl. Sci. U.S.A. 82: 3707-371 1 Grant, \l\'. S. (1977). High intertidal community organization on a rocky headland in Maine, USA Mar. Biol. 44: 15-25 LITERATURE CITED Manic, L. A.,Pringle, J. D. (1978). Pottery, a subshate for algal cullure. Br. Phycol. J. 13: 25-33 Baardseth, E. (1970). Synopsis of biological data on knobbed Hardwick-Witman, M. N., Mathieson, A. C. (1983). Intertidal wrack Ascophyllum nodosum (Linnaeus) Le Jolis. FAO macroalyae and macroinvertebrates: seasonal and spatial Fish. Synopsis. No. 38, Rev. 1 abundance patterns along an estuarine gradient. Estuar. Ballantine, W. J. (1961). A biologically-defined exposure scale coast. Shelf Sci. 16: 113-129 for the comparative description of rocky shores, Fld Stud. Hardy, F. G., Moss. B. L. (1979). Attachment and development l: 1-19 of the zygotes of Pelvetia canaliculata (L.) Dcne. et Thur. Bergeron, P., Bourget, E. (1986). Shore topography and spatial (Phaeophyceae, ). Phycologia 18: 203-212 partitioning ol crevice refuges by sessile epibenthos in an Harger, J. R. E. (1970). The effect of wave impact on some ice disturbed environment. Mar. Ecol. Prog. Ser. 28: aspects of lhe biology of sea mussels. Veliger 12: 401-414 129-145 Jackson, G. A., Winant, C. D. (1983). Effect of a kelp forest on Boney. A. D. (1966). A biology of marine algae. Hutchinson coastal currents. Cont. Shelf Res. 2: 75-80 Educational Ltd. London Jones, W. E.. Demetropoulos, A. (1968). Exposure to wave Carstens, T. (1968). Wave forces on boundries and submerged action: measurements of an importanl ecological parame- bodies. Sarsia 34: 37-60 ter on rocky shores on Anglesey. J. exp. mar. Biol. Ecol. 2: Charters, A. C., Neushul, M,, Coon, D. A. (1972). Effects of 46-63 water nlotion on algal spore attachment. Proc. In!. Sea- Josselyn. M. N., Mathieson, A. C. (1978). Contribution ol weed Syn~p.7: 243-247 receptacles from the Iucoid Ascoj?hylltml nodosum to the Coe, W. R. (1956). Fluctuations in populations of liltoral detrital pool ill a north lemperate esluary. 1: marine invertebrates. J. mar. RRS.IG: 212-232 258-26 1 Connell, J. H. (1961). Effects of competition. predation by Thals Keser, M,, Larson. B. R. (1984). Colonization and growth of lapillus, and other factors on natural populations of the AscophyUunl nodosum (Phaeophyta) in Maine. J. Phycol. barnacle Balanus balanoides. Ecol. Monogr. 31: 61-104 20: 83-87 Connell, J. H. (1972). C:ommunity interactions on marine Keser. M,. Vadas, R. L., Larson, B. R. (1981). Reglowth of rocky intertidal shores. Ann. Rev. Ecol. Syst. 3: 169-192 Ascophyllrtn? r~odosumand Fucus vesjculosus under vari- Cousens, R. (1985). Frond slze dislribulions and the effects of ous harvesting regimes in Majne, USA. Botanica mar. 24: the algal canopy on the behaviour of Ascophyllum 29 38 nodosum (L.) Le Jolls. J. exp. mar. Biol. Ecol. 92: 231-249 Koehl, M. A. R. (1977). Effects of sea anemones on the non, Cousens, R. (1986). Quantitative reproduction and reproduc- forces they encounler. J. exp. Biol. 69: 87-105 tive effort by stands of the brown alga Ascophyllum Lewis, J. R. (1964). The ecology of rocky shores. English nodosun? (L.) Le Jolis in southeastern Canada. Estuar. Universities Press, London coast. Shelf Sci. 22: 495-507 Lewis, J. R. (1968). Water movements and their role m rocky Dalby, D. H., Cowell, E. B., Syratt, W. J., Crothers, J. H. (1978). shore ecology. Sania 34: 13-36 An exposure scale for marine shores in western . J. Lewis. J. R. (1976). The role of physical and biological factors mar. biol. Ass. U. K. 58: 975-996 in the distribution and stability of rocky shore com- Dayton. P. I<.(1971). Competition, disturbance, and commun- munities. In: Keegan. B. F., Deidigh. P. 0.. Boaden, P. J. ity organization: the provision and subsequent utilization (eds.) Biology of benthic organisms. Pergamon Press, New of space in a rocky intertidal community. Ecol. Monogr. 41. York, p. 417-424 351-389 Lewis, J. R. (1980). Objectives in littoral ecology - a personal Dennv. M. W. (1985a). Water motion. In: Littler, M. M.,LitUer, viewpoint. h: Price. J. J., he,D. E. G., Farnham, W. F. D.'S. (eds.) Handbook of phycological methods IV: ecologi- (eds.) The shore environment. Vol. l. Academic hess, cal field methods: macroalqae.- Cambridge University New York, p. 1-18 Press, New York, p. 7-32 Lubchenco, J. (1978). Plant species diversity in a marine Denny. M. W. (1985b). Wave forces on intertidal organisms: a intertidal community: importance of herbivore food prefer- case study. Limnol. Oceanogr. 30: 1171-1187 ence and algal competitive abilities. Am. Nat. 112: 23-39 Mar. Ecol. Prog. Ser. 61: 263-272, 1990

Lubchenco, J. (1980). Algal zonation in the New England availability, and spatially variable patterns of succession. rocky intertidal conununity: an experimental analysis. Ecology 65: 1918-1935 Ecology 61: 333-344 Sundene, 0. (1973). Growth and reproduction in Ascophyllum Lubchenco. J. (1983). Litlorina and Fucus: effects of her- nodosum (Phaeophyceae). Norw. J. Bot. 20; 249-255 bivores, substratum heterogeneity, and plant escapes dur- Tegner, M. J. (1986). Varjations in surface currents and ing succession. Ecology 64: 11 16-1123 recruitment success. In: Thompson. M,, Sarojini, R., Lubchenco. J. (1986). Relative imporlance of competition and Nagabhushanam, R. (eds.) Biology of benthic marine predation: early colonization by in New Eng- organisms: techniques and methods as applied to the land. In: Diamond, J..Case, T. J. (eds.) Community ecol- Indian Ocean. A. A. Balkema, Rotterdam, p. 323-329 ogy. Harper and Row, New York, p. 537-555 Topinka, J. A., Tucker, L., Korjelf, W. (1981). The dislribution Mathieson, A. C., Tveter, E., Daly, M., Howard, J. (1977). of fucoid macroalqal biomass along central coastal Maine. Marine algal ecology m a New Hampshire tidal rapid. Botanica mar. 24: 311-319 Bota~licamar. 20: 277-290 Underwood, A. J. (1986). Physical factors and biological inter- McEachreon, J. C. T., Thomas, M. L. H. (1987). Attachment actions: the necessity and nature of ecological expen- strength of AscophyUum nodosum (L.) Le Jolis and expo- ments. In: Moore. P. G.. Seed, R. (eds.) The ecology of sure to wave action. Botanica mar. 30: 217-222 rocky coasts. Colunlbia Univ. Press, New York, p. 372-390 Moss, E. (1975). Attachment of zygotes and germlings of Underwood, A. J., Denley, E. J. (1984). Paradigms, explana- Ascophyllum nodosun] (L.) Le Jol. (Phaeophyceae, Fuca- tions, and generahations in models for the structure of les). Phycologia 14: 75-80 intertidal conlnlunities on rocky shores. In: Strong, D. R.. Moss, B. (1981). Attaching rnechanisms of zygotes and Simberloff, D., Abele, L. G., Thistle, A. B. (eds.) Ecological embryos of the Fucales. Proc. Int. Seaweed Symp. 8: communities: conceptual issues and the evidence. Prince- 217-124 ton University Press, Princeton, p. 151-180 Norton, T. A. (1983). The resistance to dislodgement of Sar- Vadas, R. L. (1972). Marine algae. In. Third annual report, gassum muticurn germlings under defined hydrodynarnic Maine Yankee Alon~~cPower company, Augusta. Maine, conditions. J. mar. bjol. Ass. U. K. 63: 181-193 p. 250-310 Oltmanns, F. (1889). Beitrage zurKenntnis der Fucaceen. Bibl. Vadas. R. L. (1979). Seaweeds: an overview; ecological and bot. 14: 1-94 economic importance. Experjentia 35: 429432 Ortega, S. (1981). Env~ronmental stress, competition and Vadas, R. L., Keser. M,, Larson, B. (1978). Effects of reduced dominance of Crasstostrea virginica near Beaufort, North temperatures on previously stressed populations of an inter- Carolina, USA. Mar. Biol. 62: 47-56 tidal alga. In: Thorpe, J. H.,Gibbons, J. W. (eds.) Energy and Palumbi, S. R. (1984).Measuring intertidal wave forces. J. exp. environmenlal stress in aquatic systems. DOE Synlposium mar. Biol. Ecol. 81: 171-179 Series (Conf-771114, NTIS), Augusta, Georgia, p. 434451 Patterson, B. P. (1986). Ecological relationships between Vadas, R. L., Keser, M., Larson, B., Grant, W. S. (1977). germlings of Ascopt~yllumnodosum (L.) Le Jol. and the Influence of Littorind littorea on algal zonation. Proc. Int. herbivorous snail Littorina Littorea L. M. S, thesis, Univer- Seaweed Symp. 12: 84 sity of Maine, Orono Vadas, R. L., Keser. M,,Rusanowski, P. C. (1976). Influence of Prlntz, H. (1956). Recuperation and recolonization in thermal loading on the ecology of intertidal algae. In: Esch, Ascophyllum. Proc. Int. Seaweed Symp. 2: 194-197 C. W., McFarlane, R. W. (eds.) Thermal ecology 11. ERDA Reed, D. C., Laur, D. R., Ebeliny. A. W. (1988). Variation in Symposium Series (Conf-750425, NTIS), Augusta, Geor- algal d~spersaland recruitment: the importance of episodic gia, p. 202-212 events. Ecol. Monogr. 58: 321-335 Vadas, R. L., Miller, S. L., Bolis, C. M,,Bacon. L., Wright. W. SAS Institute Inc. (1982). SAS user's guide: st-atistics. SAS (1982). Population dynamics of Ascophyllum nodosumt Institute Inc.. Cary, NC factors influencing recruitment and germlings. 1st Int. Schonbeck, M. W., Norton, T. A. (1980). Factors controlling Phycol. Cong., Memorial University St. Johns, Newfound- the lower limits of fucoid algae on the shore. .l. exp. mar. land, Canada, p. 52 Biol. Ecol, 43: 131-150 Vadas, R. L., Wright, W. A. (1986). Recruitment, growth and Schwenke, H. (1971). Water overnent, 2. Plants. In: Kinne, 0. management of Ascophyllum nodosunl. Actas I1 Congr. (ed.) Marine ecology. Vol. l. Part 2. \Yiley, London, Algas mar. chil. 2: 101-113 p. 1091-1 121 Vahl, 0. (1982). Long-term variations in recruitment of the Sebens, K. P. (1982). Asexual reproduction in .4mthopleura Iceland scallop Cl~lamysislandica irom Northern Norway. elegantissjma (Anthozoa: Actiniaria): seasonality and spa- Neth. J. Sea Res. 16: 80-87 tial extent of clones. Ecology 63: 434-444 Velm~irov. B. (1983). Succession in a kelp bed ecosystem; Sebens, K. P,, Lewis, J. R. (1985). Rare events and population clearing of primary substrate by wave-induced kelp structure oi the barnacle Sen~jbalanus cariosus [Pallas, sweeping. Oceanologica Acta No SP p. 201-206 1788). J. exp. mar. Biol. Ecol. 87: 55-65 Vogel. S. (1981). Ljfe in moving fluids, the physical hjology of Sheader, A., kloss, B. (3975). Effects of Light and temperature flow. Princeton Univ. Press, Princeton on gernlination and growth of Ascophyllu~m nodosum (L.) Watson, D. C., Norton, T. A. (1985). Dietary prelererlces of the Le Jol. Esluar. coast. mar. Sci. 3: 125-132 cotnmon periwinkle, Littorina littorea (L.)J. exp. mar. Blol. Sokal, R. R., Rohlf, F. J. (1981). Biometry, 2nd edn. W. H. Ecol. 88: 193-211 Freeman and Co.. New York White, A. S. (198.5). Presettlement regeneration patterns in a Sousa, W. P. (1984), intertidal mosaics: patch size, propagule southwestern ponderosa pine stand. Ecology 66;: 589-594

This article was pre.sented by Professor M Neusl~ul, Santd ~Manirscript first received: Septen~ber26, 1988 Bsrbara, Ct~lifol.nia, USA Revised version accepted: December 12, 1989